Structural Prominence, Network Density, and Performance in Supply Networks: A Complex Adaptive Systems Perspective

Supply networks are complex interorganizational systems in which structural prominence may have different associations with operational and financial outcomes. We examine 169 firm-year observations from 38 focal firms in 10 industries. The primary three-knot restricted cubic-spline models include industry and year indicators, firm-clustered small-sample inference, simultaneous derivative bands, revision-defined supplementary regional average-slope analyses, direct low-versus-high slope contrasts, and wild-cluster bootstrap checks. Some regional average slopes are positive, but none of the direct regional contrasts is significant, no centrality relationship is globally nonlinear, and none of the overall centrality tests reaches 0.05 under wild-cluster inference. The pooled inventory and OPR estimates shrink markedly in Mundlak/CRE and firm fixed-effects models; the ROA average derivative remains similar in magnitude but the joint spline terms are not significant. The evidence therefore does not establish that centrality associations differ across regions or that the underlying functions contain distinct regional transitions. Density has a positive linear inventory coefficient under the small-cluster test but not the wild-cluster test, and H3 is not supported as an overall density–ROA association. Because the retained analytic file lacks node identifiers, component membership, shipment frequency, and volume, focal-network overlap, directed or harmonic closeness, and alternative edge definitions cannot be reconstructed.